Optimizing a machine learning based glioma grading system using multi-parametric MRI histogram and texture features Optimizing a machine learning based glioma grading system using multi-parametric MRI histogram and texture features

The methylation status of oxygen 6-methylguanine-DNA methyltransferase (MGMT) promoter has been associated with treatment response in glioblastoma(GBM). Using pre-operative MRI techniques to predict MGMT promoter methylation status remains inconclusive. In this study, we investigated the value of features from structural and advanced imagings in predicting the methylation of MGMT promoter in primary glioblastoma patients. Ninety-two pathologically confirmed primary glioblastoma patients underwent preoperative structural MR imagings and the efficacy of structural image features were qualitatively analyzed using Fisher’s exact test. In addition, 77 of the 92 patients underwent additional advanced MRI scans including diffusion-weighted (DWI) and 3-diminsional pseudo-continuous arterial spin labeling (3D pCASL) imaging. Apparent diffusion coefficient (ADC) and relative cerebral blood flow (rCBF) values within the manually drawn region-of-interest (ROI) were calculated and compared using independent sample t test for their efficacies in predicting MGMT promoter methylation. Receiver operating characteristic curve (ROC) analysis was used to investigate the predicting efficacy with the area under the curve (AUC) and cross validations. Multiple-variable logistic regression model was employed to evaluate the predicting performance of multiple variables. MGMT promoter methylation was associated with tumor location and necrosis (P < 0.05). Significantly increased ADC value (P < 0.001) and decreased rCBF (P < 0.001) were associated with MGMT promoter methylation in primary glioblastoma. The ADC achieved the better predicting efficacy than rCBF (ADC: AUC, 0.860; sensitivity, 81.1%; specificity, 82.5%; vs rCBF: AUC, 0.835; sensitivity, 75.0%; specificity, 78.4%; P = 0.032). The combination of tumor location, necrosis, ADC and rCBF resulted in the highest AUC of 0.914. ADC and rCBF are promising imaging biomarkers in clinical routine to predict the MGMT promoter methylation in primary glioblastoma patients.

The preoperative grading of gliomas, which is critical for guiding therapeutic strategies, remains unsatisfactory. We aimed to retrospectively assess the efficacy of intravoxel incoherent motion (IVIM) diffusion-weighted imaging (DWI) in the grading of gliomas. Forty-two newly diagnosed glioma patients underwent conventional MR imaging, DWI, and contrast-enhanced MR imaging. Parameters of apparent diffusion coefficient (ADC), slow diffusion coefficient (D), fast diffusion coefficient (D*), and fraction of fast ADC (f) were generated. They were tested for differences between low- and high-grade gliomas based on one-way ANOVA. Receiver-operating characteristic (ROC) analyses were conducted to determine the optimal thresholds as well as the sensitivity and specificity for grading. ADC, D, and f were higher in the low-grade gliomas, whereas D* tended to be lower (all P<0.05). The AUC, sensitivity, specificity and the cutoff value, respectively, for differentiating low- from high-grade gliomas for ADC, D and f, and differentiating high- from low-grade gliomas for D* were as follows: ADC, 0.926, 100%, 82.8%, and 0.7 × 10(-3) mm(2)/sec; D, 0.942, 92.3%, 86.2%, and 0.623 × 10(-3) mm(2)/sec; f, 0.902, 92.3%, 86.2%, and 35.3%; D*, 0.798, 79.3%, 84.6%, and 0.303 × 10(-3) mm(2)/sec. The IVIM DWI demonstrates efficacy in differentiating the low- from high-grade gliomas.

Background: Based on conventional MRI images, it is difficult to differentiatepseudoprogression from true progressionin GBM patients after standard treatment, which isa critical issue associated with survival. The aim of this study was to evaluate the diagnostic performance of machine learning using radiomics modelfrom T₁-weighted contrast enhanced imaging(T₁CE) in differentiating pseudoprogression from true progression after standard treatment for GBM. Methods: Seventy-sevenGBM patients, including 51 with true progression and 26 with pseudoprogression,who underwent standard treatment and T₁CE, were retrospectively enrolled.Clinical information, including sex, age, KPS score, resection extent, neurological deficit and mean radiation dose, were also recorded collected for each patient. The whole tumor enhancementwas manually drawn on the T₁CE image, and a total of texture 9675 features were extracted and fed to a two-step feature selection scheme. A random forest (RF) classifier was trained to separate the patients by their outcomes.The diagnostic efficacies of the radiomics modeland radiologist assessment were further compared by using theaccuracy (ACC), sensitivity and specificity. Results: No clinical features showed statistically significant differences between true progression and pseudoprogression.The radiomic classifier demonstrated ACC, sensitivity, and specificity of 72.78%(95% confidence interval [CI]: 0.45,0.91), 78.36%(95%CI: 0.56,1.00) and 61.33%(95%CI: 0.20,0.82).The accuracy, sensitivity and specificity of three radiologists’ assessment were66.23%(95% CI: 0.55,0.76), 61.50%(95% CI: 0.43,0.78) and 68.62%(95% CI: 0.55,0.80); 55.84%(95% CI: 0.45,0.66),69.25%(95% CI: 0.50,0.84) and 49.13%(95% CI: 0.36,0.62); 55.84%(95% CI: 0.45,0.66), 69.23%(95% CI: 0.50,0.84) and 47.06%(95% CI: 0.34,0.61), respectively. Conclusion: T₁CE–based radiomics showed better classification performance compared with radiologists’ assessment.The radiomics modelwas promising in differentiating pseudoprogression from true progression.

Aims: To investigate the white matter (WM) integrity and hippocampal functional connectivity (FC) in type 2 diabetes mellitus (T2DM) patients without mild cognitive impairment (MCI) by using diffusion tensor imaging (DTI) and resting-state functional magnetic resonance imaging (rs-fMRI), respectively. Methods: Twelve T2DM patients without MCI and 24 age, sex and education matched healthy controls (HC) were recruited. DTI and rs-fMRI data were subsequently acquired on a 3.0T MR scanner. Tract-based spatial statistics (TBSS) combining region of interests (ROIs) analysis was used to investigate the alterations of DTI metrics (fractional anisotropy (FA), mean diffusivity (MD), λ₁ and λ₂₃) and FC measurement was performed to calculate hippocampal FC with other brain regions. Cognitive function was evaluated by using Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA). Brain volumes were also evaluated among these participants. Results: There were no difference of MMSE and MoCA scores between two groups. Neither whole brain nor regional brain volume decrease was revealed in T2DM patients without MCI. DTI analysis revealed extensive WM disruptions, especially in the body of corpus callosum (CC). Significant decreases of hippocampal FC with certain brain structures were revealed, especially with the bilateral frontal cortex. Furthermore, the decreased FA in left posterior thalamic radiation (PTR) and increased MD in the splenium of CC were closely related with the decreased hippocampal FC to caudate nucleus and frontal cortex. Conclusions: T2DM patients without MCI showed extensive WM disruptions and abnormal hippocampal FC. Moreover, the WM disruptions and abnormal hippocampal FC were closely associated. Highlights - T2DM patients without MCI demonstrated no obvious brain volume decrease. - Extensive white matter disruptions, especially within the body of corpus callosum, were revealed with DTI analysis among the T2DM patients. - Despite no MCI in T2DM patients, decreased functional connectivity between hippocampal region and some critical brain regions were detected. - The alterations in hippocampal functional connectivity were closely associated with those of the white matter structures in T2DM patients. This trial was registered to ClinicalTrials.gov (NCT02420470, https://www.clinicaltrials.gov/).